Efficient monolithic optocoupler device

This application is a continuation of U.S. application Ser. No. 16/563,463, filed Sep. 6, 2019, which claims the benefit of U.S. Provisional Application No. 62/728,463, filed Sep. 7, 2018, Provisional Application No. 62/800,805, filed Feb. 4, 2019, Provisional Application No. 62/800,809, filed Feb. 4, 2019, and Provisional Application No. 62/824,037, filed Mar. 26, 2019. The entire contents of those applications are incorporated by reference.

DC-DC conversion is a very common element of a variety of circuits. Very efficient step down (“buck”) converters and step up (“boost”) converters are available, which use a switching technique to regulate the voltage. However, there are several drawbacks to switching DC-DC converters. Firstly, they require additional components such as inductors and capacitors, which increase the footprint and cost of the solution. Also, high-frequency switching introduces voltage ripple, which can create problems in noise-sensitive applications and requires additional design complexity to suppress. Also, switching converters are susceptible to electromagnetic interference effects which can create a variety of problems in electronic circuits.

Alternative devices for DC-DC conversion are linear voltage regulators, which are less complex than switching converters, require fewer external components and there is no noise generated by switching. Therefore, these devices can be low cost, insensitive to EMI and a very compact alternative to switching converters. However, there are two main drawbacks with linear voltage regulators. The first is that they only step down voltage, severely limiting the cases where the devices can be employed. The input voltage is required to be greater than the output voltage by an amount known as the dropout voltage, which in low dropout (LDO) linear regulators can be as low as 50-100 mV, or as high as 2V in conventional linear voltage regulators. The second drawback is that their efficiency is generally lower, especially when the difference between input and output voltages is large. In battery operated systems this equates to increased battery drainage and significant amounts of waste heat generation.

In this invention, a device is provided that operates in the manner of a linear voltage regulator, but with the functionality of a switching converter. This concept enables DC voltage up-conversion with no switching, more efficient step down of large voltage steps, and requires no expensive and bulky additional components. At the heart of the invention is an optocoupler device which transfers power between light emitting and photovoltaic devices.

In describing the illustrative, non-limiting embodiments of the invention illustrated in the drawings, specific terminology will be resorted to for the sake of clarity. However, the invention is not intended to be limited to the specific terms so selected, and it is to be understood that each specific term includes all technical equivalents that operate in similar manner to accomplish a similar purpose. Several embodiments of the invention are described for illustrative purposes, it being understood that the invention may be embodied in other forms not specifically shown in the drawings.

A standard linear voltage regulator operates by taking an unregulated input, V, and producing a regulated output, V. The regulation is achieved by a variable voltage drop across a transistor connected to the output. The transistor is controlled by an error amplifier which compares the output voltage to a voltage reference, V. In this circuit, Vis always lower than V. Very low dropout voltages can be achieved by modifying this basic circuit to use a common emitter output stage as described, for example, in Horowitz & Hill

Turning to the drawings,shows one illustrative, non-limiting embodiment of the improved linear voltage regulator circuitof the invention. The circuitincludes a transistor, an amplifier, and an optocoupler section. The optocoupler sectionhas a light emitting sectioncomprising an array of LEDs, and a photovoltaic sectioncomprising an array of photovoltaic devices. The photovoltaic devicecan be any suitable devices, for example a photosensor, photodetector, solar cell, or a semiconductor junction (e.g., p-n junction) that has an anode and cathode contact and absorbs external photons above the bandgap of the lowest bandgap material in the device, which can be extracted as electrical current in an external circuit.

The transistoris not connected to the output, as in a conventional voltage regulator, but rather to one or more light emitting diodesconnected in a combination of series or parallel connected strings. The circuit has a transistorconnected to an input terminalwith an unregulated voltage. The circuit is shown with a npn bipolar junction transistor, but analogous operation could be achieved using other types of transistor, such as a MOSFET. The function of the circuit is to produce a regulated voltage output at the output terminal.

A third terminalprovides a voltage reference, which may be an external voltage source, a suitable diode or other similar component for providing a voltage reference. The voltage drop across transistoris controlled by an error amplifierwhich compares the voltage at the output terminalthrough a potential divider, formed by resistorsand, with the voltage reference at the reference terminal. The error amplifier, when in a feedback loop, generates a voltage which stabilizes the output voltage and minimizes the difference between the reference and the feedback voltage.

Thus, in the embodiment shown, the amplifier has a first input connected to the voltage reference, and a second input connected to the output terminalthrough the potential divider. Specifically, the first and second resistors,of the voltage divider are connected in series with the second input of the amplifierconnected therebetween. The first resistorhas a first end that is connected to the output terminaland a second end that is connected to a first end of the second resistor. The second end of the second resistoris connected to ground. The second input to the amplifieris connected to the second end of the first resistorand the first end of the second resistor.

The potential divider provides the feedback voltage which is compared to the reference voltage in the amplifier. The feedback voltage is a fraction of the output voltage, controlled by the resistors' values, thereby controlling the regulated voltage output. By changing the values of these resistors, the voltage regulator can be modified to produce different output voltages (some designs of regulator ICs require separate resistors to be connected off the chip, others have the resistors built-in to the IC). Using a variable resistor as one or both of the resistors in the divider is an easy way to make a variable voltage regulator.

The output from the amplifieris connected to the base of the transistor, and the collector of the transistor is connected to the input terminal. The connection between the voltage referenceand the input terminalis optional and depends on the type of voltage reference being used. If the voltage reference is a diode then the diode is connected to the inputthrough a resistor. If an external reference (e.g. power supply) is used as a reference, terminaldoes not have to be connected to the input terminal.

In addition, the emitter of the transistoris connected to a light emitting section. The light emitting sectionis an array of one or more light emitting diodes (LEDs)which are connected in any combination of series and parallel interconnections. The light emitter sectionexample shown inhas four LED devicesconnected in parallel. The LEDstransmit photons to a photovoltaic section, which is an array of one or more photovoltaic (PV) devicesconnected in a combination of series or parallel strings. The direction of photon transfer follows the arrow shown by, and therefore the bandgap energy of the PV devices is required to be equal to or less than the energy of photons emitted by the LED devices—but only the narrowest bandgap part of the PV device is needed to have a bandgap equal to or less than the emission energy of the LED. For example, the PV device may comprise a heterojunction having a wide bandgap n region and a narrower bandgap p region. A PV deviceis aligned with a respective LEDto receive and detect light emitted from the LED. The LEDsemit a light in response to a signal received from the emitter of the transistor.

The transistorprovides a fraction of the input voltage to the LEDs, controlled by the amplifier feedback loop. The intensity and wavelength of the light from each LEDis substantially the same, although series resistance and real-world variations in LEDs make that an approximation. The fraction of the input voltagenot passed to the LEDarray is given up as heat in the transistor. The output from the last PV devicein the series connects to the output terminal.

The PV sectionhas four PV devicesconnected in series. The LED and PV devices may be separate components, or monolithic devices with high impedance between the terminals of the LED and PV sections. The PV section is connected to the output terminal. There is no requirement for the number and performance attributes of the LED and PV devices in their respective strings to be identical. For example, it is possible that one large LED devicecould pass photons to several smaller PV devices, or conversely, one large PV devicecould absorb photons from several smaller LEDs. By using different combinations of series and parallel connections of the LEDand PV devices, step-up or step-down voltage conversion is possible with this scheme.

For simplicity, no additional circuit elements for overcurrent protection or stability are shown in. A variety of capacitors, resistors and transistors may be added to the circuit to improve stability and protect against short circuits, but do not govern the basic operating principle of the regulator device. The voltage produced across the terminals of the PV sectionprovides a voltage boost at the output of the optocoupler sectionrelative to the input of the optocoupler section. For example, the voltage across the PV sectionis greater by a factor of roughly 4× relative to the voltage across the LED section when the bandgap energies of the LEDand PV devicesare very similar.

More specifically, photovoltaic devices in series add their voltages, analogous to connecting batteries in series. Thus, for example, consider 4 LEDs in parallel and 4 PV devices in series where the PV and LED bandgaps are very similar and the LED luminescence is coupled to the PV devices. Neglecting optical and electrical losses for the simplicity of this example, the voltage produced by the PV array will be roughly 4× greater than the voltage input to the LED array, whereas the current produced by the PV array will be roughly 4× smaller than the current input to the LED array, thereby ensuring the output power from the PV array cannot exceed the power input to the LED array. In practice, electrical and optical losses will reduce the voltage boost to less than 4× and reduce the current output of the PV array to less than ¼ of the LED input, resulting in significantly less than unity power transfer efficiency of the optocoupler, where power transfer efficiency has been defined as the output power from the PV array as a fraction of the input power to the LED array.

In general, the design of the LED sectionand PV sectionare designed such that the output voltage is close to the maximum power voltage of the PV sectionunder typical operating voltage ranges, which provides the highest efficiency of electrical power out of the PV sectionper incident light power.

The linear regulatorhas a feedback loop that regulates the voltage. The voltage referenceis compared to the voltage across the potential divider. The amplifierminimizes the difference between the reference voltageand the feedback voltage, by controlling the voltage drop across the transistor. The optocoupler section or deviceprovides additional voltage reduction or voltage boost between the input and output sides, which enables either voltage upconversion, or more efficient voltage down conversion.

In, all the power is transferred between the light emitter and the PV devices. The efficiency of the optocoupler is less than unity, so if there were no voltage boosting or voltage reduction going on in the optocoupler section, therefore comparable to a conventional voltage regulator, the efficiency would be lower using the optocoupler. However, when stepping down large voltage steps, reducing voltage in the optocoupler section enables a higher efficiency than a conventional linear regulator where the entire voltage difference is dissipated as heat in the transistor. Furthermore, by employing an optocoupler which enables voltage boosting, DC-DC upconversion can be achieved using a device which retains the advantages of a linear regulator in simplicity, low EMI and low voltage ripple.

In, is a low dropout voltage regulator with an optocoupler section. It is a similar configuration toto provide to connect LEDs and PV devices into voltage regulator circuits which are low dropout regulator circuits. Here, the circuit has an input terminalwith an unregulated voltage and produces a regulated voltage at the output terminal. As in, a third terminalprovides a voltage reference. The voltage drop between the input terminaland the light emitting sectionis now across a transistorin a common emitter configuration, which in this example uses a pnp bipolar junction transistor. The emitter of the transistoris connected to the input terminal, and the collector is connected to the LED section. The base of the transistoris connected to the emitter of a second transistorhaving a collector connected to ground. The second transistorhas a base that is connected to and controlled by an error amplifier, which compares the voltage at the output terminalthrough a potential divider formed by resistorsand, with the voltage reference at the reference terminal, as in. It provides lower parasitic voltage loss in the transistor, so that the output voltage can be closer to the input voltage.

Also analogous to the example in, the transistoris connected to a light emitting section, which is an array of one or more LEDsconnected in any combination of series and parallel connections. The LEDstransmit photons to a photovoltaic section, having of one or more photovoltaic (PV) devicesconnected in a combination of series or parallel strings. The direction of photon transfer follows the arrow shown by. As in, for simplicity, no additional circuit elements for overcurrent protection or stability are shown. A variety of capacitors, resistors and transistors may be added to the circuit to improve stability and protect against short circuits, but are not required to understand the basic operating principles of the regulator device.

For applications where the output voltage of the PV sectionis greater than the input to the light emitting section, achieved by building voltage using series connections of PV devices, linear regulators with negative dropout voltage can be realized. In other words, this represents a linear regulator device capable of DC voltage boost conversion. Alternatively, when the output voltage of the PV sectionis significantly less than the input voltage to the light emitting section, for example by using multiple LED devices in series, a greater efficiency for step-down conversion of large voltage steps can be achieved, compared to conventional linear regulators. Using conventional regulators for large voltage reduction, all the voltage change would be governed by the voltage drop across a transistor, leading to a significant amount of wasted power and producing heat. In the circuits of, a fraction of the voltage drop can be achieved using the photon transfer process, thereby raising the efficiency overall.

Galvanic Isolation

shows an alternative example, non-limiting design of the regulator described incan be used to create a linear voltage regulator device which has a high degree of galvanic isolation between the input and output sides of the device. The circuit has an input terminalwith an unregulated voltage and the function of the circuit is to produce a regulated voltage output at the output terminal. A third terminalprovides a voltage reference. Instead of a transistor, a phototransistoris used to create a controlled voltage drop. The voltage drop across the phototransistoris controlled by a light source such as a control light emitting deviceas part of an optocoupler architecture. The control LEDis connected to the output of an error amplifierwhich compares the voltage at the output terminalthrough a potential divider, formed by resistorsand, with the voltage reference at the reference terminal, as in.

The phototransistorhas a collector that is connected to the input. The emitter of the phototransistoris connected to a light emitting section, which transmits photons to a photovoltaic section, entirely analogous to the schemes shown in. Thus, in response to light received by the control LED, the phototransistorprovides a control signal to the LEDs. The direction of photon transfer follows the arrow shown by. In this embodiment, the inputand outputterminals are completely galvanically isolated, with the only current path between the two sides being via photon transfer. Furthermore, a similar circuit tocan be constructed by replacing the phototransistor output stagewith a common emitter output stagecontrolled by a pnp phototransistor, as shown in.

In, the only way current can flow from the input to the output side is via photons (especially if the PV and LEDs are discrete components), which restricts the current level which can flow. Thus, it provides galvanic isolation, which can be a useful safety device preventing potentially dangerous electric shocks. Another important use for galvanic isolation is when the input and output sides of the circuit have different ground potentials.

High Efficiency Version

shows another alternative example non-limiting design of the regulator used to create DC voltage conversion with high efficiency. Here, the circuit has an input terminalwith an unregulated voltage and produces a regulated voltage at the output terminal). As in, a third terminalprovides a voltage reference, and the voltage drop between the input terminaland the light emitting sectionis across a transistorin a common emitter configuration. The base of the transistoris connected to the emitter of a second transistor. The second transistoris controlled by an error amplifierwhich compares the voltage at the output terminalthrough a potential divider, formed by resistorsand, with the voltage reference at the reference terminal, as in.

Analogous to the example in, the transistoris connected to a light emitting section. The LEDs transmit photons to a photovoltaic section, having one or more photovoltaic (PV) devicesconnected in a combination of series or parallel strings. In the example shown in, three LEDsconnected in series are optically coupled to three PV devicesconnected in parallel. When the bandgap energies of the LED and PV devices are similar, this configuration results in a roughly 3× reduction of the voltage across the PV section relative to the voltage across the LED section. The direction of photon transfer follows the arrow shown by.

Unlike, in the embodiment inthe input terminalis also connected directly to the PV section. By connecting one terminal of the photovoltaic string to the input, the photovoltaic action adds or subtracts to the voltage supplied to the load. In this configuration, only a fraction of the power supplied at the input is used to drive the light emitting section, and the overall efficiency can be significantly greater than if all the power from the input were routed through the light emitter section.

In the example in, when the input side voltage exceeds the regulated output voltage, some of the unwanted power is dissipated in the photovoltaic section. To minimize this, another similar embodiment of the idea places a conventional linear regulatorin parallel with the circuit, as shown in. The input terminal of the linear regulatoris connected to the input terminalof the device and the output terminal of the linear regulatoris connected the output terminalof the device. All other aspects of the circuit are identical to the one shown in. When the input voltage at the input terminalexceeds the input voltage threshold of the standard linear regulator, the power is largely diverted through the conventional regulator circuit. This can either be controlled by a switch, for example a single pole double throw switch, or by ensuring the output voltage from the conventional regulator is set to be slightly greater than the main voltage regulator.

The embodiments shown inenable an improvement in efficiency compared to the embodiments in, as the power transfer efficiency of the optocoupler section can be significantly less than unity. In, part of the power passes directly from the inputto the output, so that only part of the power passing from the input to the output is transferred through the optocoupler region, and therefore the impact on overall efficiency due to the less than unity optocoupler power transfer efficiency is reduced.

Monolithic Optocoupler Design

Optocouplers are a common device in a variety of applications, and usually discrete devices for photon generation and photon absorption. Separate components increases cost and makes efficient optical coupling difficult to achieve. Generally, optocoupler devices are used for signal transmission and electrical isolation, not for efficient power transfer. Monolithic optical emitter/detector devices have been demonstrated which improve the coupling efficiency of photons. For example, U.S. Pat. No. 4,275,404 to Cassiday et al. devised a device where an LED emitter is positioned in between two photodiode devices, all made from the same epitaxially grown layers, on an insulating substrate. The emission from the side of the LED section is coupled into the photodetector sections, creating an opto-isolator. Vertical optical connections have also been demonstrated, for example by U.S. Pat. No. 5,753,928 to Krause et al. Here, a single emitter region is stacked monolithically with a detector region to produce a monolithic optical emitter-detector. Voltage multiplication has also been demonstrated in devices such as the Toshiba TLP590B. Here a single discrete LED is optically coupled to a series-connected silicon photodiode array to produce a greater voltage output. However, the power transfer efficiency of this device is very low due to the highly inefficient production, transfer and electrical conversion of photons in the device.

To achieve high power transfer efficiency, a device is provided where the photon generation and absorption are performed by different regions of the same monolithic device, separated by transparent, highly resistive, monolithic isolation layers. Step-up or step-down voltage conversion is possible if the light emitting or light absorbing regions are made up of multiple devices connected in series. High efficiency photon capture is enabled by using one or more high reflectivity mirrors or a sandwiched structure, where PV regions surround a light emitting region in monolithic stack.

shows a schematic drawing of an example voltage reducing optocoupler devicehaving monolithic light emitting devicesand monolithic photovoltaic devices. Here two light emitting regions, for example GaAs pn junctions, are connected to a blocking region. The blocking region is a material layer, or combination of material layers, with high transparency to the light emitting section luminescence and high electrical resistance, for example a vertical stack of one or more AlGaAs pn junctions. The blocking regionallows photons through, but blocks electrical current, i.e., has a high electrical resistance. The blocking region is connected to a photovoltaic region. The bandgap of the lowest bandgap absorber layer in the solar cell region should be equal to or less than the bandgap of the light emitting section governing the peak emission wavelength.

An optional reflectoron the light emitting section devices functions to improve the coupling efficiency of photons to the photovoltaic region. This could be an epitaxially-grown reflector such as a distributed Bragg reflector (DBR), or a separately deposited reflector such as a dielectric DBR or metal mirror. An optional reflectoron the photovoltaic region device functions to improve the absorption probability of photons emitted from the light emitting regionin the photovoltaic region.

The reflectors,, photovoltaic region, blocking region, and light emitting regionseach have linear flat top and bottom surfaces and can all have substantially the same length and width, though the light emitting regionscan be one device or multiple separate devices, as shown, so that the light emitting regionsextend substantially along the blocking regionand photovoltaic region. The bottom surface of the photovoltaic region is mounted on (i.e., coupled) to the top surface of the first reflector, and the top surface of the photovoltaic regionis coupled to the bottom surface of the blocking region. The one or more light emitting regionscan be mounted (at the bottom surface of the light emitting regions) to the top surface of the blocking region. The bottom surface of the second reflector(s)are coupled to the top surface of the light emitting regions. There can be a gap between the light emitting regions, as shown, or no gap. Thus, the elements are stacked in a vertical configuration.

In physically separate components, there are reflection losses for light escaping the LED, then again entering the photovoltaic. Furthermore, only light emitted within the escape cone of the LED can escape, the rest is reflected back into the LED by total internal reflection. In the monolithic optocoupler of the present invention, the refractive index of the LEDand PVare very similar, so the critical angle is almost 90 degrees, suppressing the escape cone limitation and reflection loss. Another benefit of a monolithic device is that the LEDand PV devicesare aligned in close proximity. The other typical loss is that 50% of the LED light is emitted in the direction away from the PV device, but the mirror serves to reflect that light back toward the PV device.

can use any suitable optocoupler device. In addition, the optocoupler devices ofcan be used in any suitable circuit configuration, such as for example, as photovoltaic output optocouplers. One application for these is for MOSFET gate driving.

However,are all examples of the structure which could be used for the optocoupler sectionof the regulatorsshown in. In that instance, the optocoupler deviceofare represented as the optocoupler devicesof. And, the light emitting deviceand PV devicesofcorrespond to the light emitting devicesand PV devicesof.

Referring to, the light emitting regionhas two terminalsand, and in this example, two light emitting devices connected in electrical series. The photovoltaic regionhas two terminalsand, and one photovoltaic device. The photovoltaic device can be centered and positioned between the two light emitting devices. When the bandgap energy of the light emitting and photovoltaic devices are similar, the input voltages between terminalsandis roughly a factor of two or more greater than the voltage between terminalsand. This could be an epitaxially-grown reflector such as a distributed Bragg reflector (DBR), or a separately deposited reflector such as a dielectric DBR or metal mirror.

The example incontains a single photovoltaic device large enough to accept photons from both devices in the light emitting regions. Alternatively, as shown in, an equivalent scheme could use two photovoltaic devicesconnected in parallel, one for each light emitting region, and with a respective blocking device. In general, a single optocoupler may have one (1) light emitting device and one (1) photovoltaic device, or any number of series or parallel connected light emitting and photovoltaic devices with any ratio of device areas.

In, having 1 LED and 1 PV per individual monolithic stack provides a basic building block to get a large number of possible ratios of voltage between LED and PV strings by connecting arbitrary numbers of them in any series/parallel combination, whereasgives a fixed 2:1 combination. However, additional metal interconnections are needed into achieve the same result as in, which may introduce more series resistance.

show an alternative example non-limiting configuration for reducing voltage. In this example, the optocoupler includes two light emitting regionswhich are connected to two blocking regionswhich are connected to a single photovoltaic region, in a vertically stacked configuration. Each light emitting regioncontains an optional reflectorto improve the coupling efficiency of photons to the photovoltaic region. As before, the bandgap of the lowest energy absorber layer in the solar cell sections should be equal to or less than the bandgap of the light emitting section governing the peak emission wavelength. The light emitting regioncontains two terminals,and, in this example, two light emitting devices connected in electrical series. The photovoltaic regioncontains two terminals,and one photovoltaic device. When the bandgap energy of the light emitting and photovoltaic devices are similar, the input voltages between terminals,is roughly a factor of two or more greater than the voltage between terminals,.

In the case that the bandgap of the light emitting regionsand the photovoltaic regionsof the devices inare equal, in principle the optocoupler can operate also as a voltage increasing device. This is achieved by operating the PV section as a light emitter and the light emitter region as a photovoltaic. However, to achieve high power transfer efficiency, it is often advantageous to optimize the layer structure to perform for either one of step-up or step-down operations.

shows an example configuration for the device as a voltage increasing device. Here a single light emitting regionis connected to a blocking region. The blocking region is connected to a photovoltaic region comprising two separate photovoltaic devices. The bandgap of the lowest bandgap absorber layer in the solar cell region should be equal to or less than the bandgap of the light emitting section governing the peak emission wavelength. An optional reflectoron the light emitting section devices functions to improve the coupling efficiency of photons to the photovoltaic region. Thus, the light emitting regionis on the bottom and the photovoltaic regionsare on top.

As shown in, the light emitting regioncontains two terminals,and, in this example, comprises a single light emitting device. The photovoltaic regioncontains two terminals,and comprises two photovoltaic devices connected in electrical series. When the bandgap energy of the light emitting and photovoltaic devices are similar, the output voltages between terminals,is roughly a factor of two greater than the voltage between terminals,. An optional reflectoron the photovoltaic region device functions to improve the absorption probability of photons emitted from the light emitting regionin the photovoltaic region. The example incontains a single light emitting device large enough to provide photons to both devices in the photovoltaic region. Alternatively, an equivalent scheme could use two light emitting devices connected in parallel, as shown in.

shows an alternative method for increasing voltage. In this example, the optocoupler includes two photovoltaic regionswhich are connected to two blocking regionswhich are connected to a single light emitting region. Each photovoltaic region contains an optional reflectorto improve the coupling efficiency of photons to the photovoltaic region.